Technology convergence and digital divides.

Transkrypt

Technology convergence
and digital divides.
A country-level evidence for the period
2000–2010
Ewa Lechman, Ph.D., Faculty of Management and Economics, Gdañsk University of
Technology
Key words: technology, convergence, ICTs, quantile convergence, clusters,
technology clubs
JEL codes: C22, O11, O50, 033
1. Digital divide—concept clarification.
The notion of digital divide is fully connected with new information and
communication technologies (ICTs). Information and Communication Technologies—ICTs, understood as means of communication, storage and retrieving all kinds of knowledge and information. In recent years a very fast adoption of ICTs in a wide set of countries has been reported.
Digital technologies are broadly considered of great importance for enhancing both social and economic development. However, new technologies
have a great ability to spread at a high pace, along with their fast adoption in
many countries, growing inequalities may appear. The unequal distribution
of ICTs was a point of interest for Schramm [1964], Sussman and Lent [1991],
and later—for example—Schiller [1996]. As proven in the works of the cited
authors, fast diffusion of new technologies is broadly considered to be
accompanied by their uneven distribution.
Growth rates showing the speed of changes in the ICTs’ field are astonishing, and the period of (for example) ten years can bring crucial changes on the
world map. If we take into account i.e. such indicators as the Internet users or
mobile cellular subscribers, the annual growth rates achieve an average
level of 50–60%1. As is widely recognized, a fast implementation of new technologies, however positive in nature, can create huge disparities in inter-country ICTs application (see Table 1). This would suggest that a fast growth
in ICTs adoption causes increasing inequalities among countries in the field.
1
Own estimates based on data derived from International Telecommunication Union data-
base.
96
ekonomia 31
Technology convergence and digital divides. A country-level evidence for the period…
Differences in the level of digitalization bring to mind the notion of “digital
divide” also called “digital gap”, “technology divide” or “technology gap”. In
recent literature, there is a multitude of ways to define the digital divide. Different authors conceptualize the digital divide differently and adopt a variety of ways to measure it. The most common definition of the digital gap is the
one presented in “Understanding digital divide” [OECD 2001], where the
term refers to the
gap between individuals, households, businesses and geographic areas at different
socio-economic levels with regard to their opportunities to access information and
communication technologies and to their use for a wide variety of activities.
The cited definition, even if very general, reveals the very nature of the problem. Whatever definition we would work out, it always will refer to differences in access to ICTs. It also refers to a kind of separation between those
who have and those who are permanently lacking access to ICTs tools. The dichotomy between “haves” and “have-nots” is revealed at the same time. The
simple notion of digital divides usually refers solely to technical access,
which from an analytical perspective is narrow. However, it is usually perceived as such—taking into account simple access to Internet and/or to other
ICTs tools.
Berlot (2003) and other authors point to the significance of such dimensions of digital divide as information technology literacy or effective usage of
ICTs. DiMaggio and Hargittai [2001] also stress the importance of ICTs usage
patterns, skills enabling to use ICTs in a proper and effective way. Devaraj
and Kohli [2003], Zhu and Kraemer [2005] point out the importance of gains
that business sector can acquire by employing ICTs—consequently they define digital (technology) gap from a strictly business perspective.
The digital divide however can be analyzed on 3 levels: country, company,
household or individual level. Dewan and Riggins [2005] distinguish three
different levels of analysis of digital divide. These are: individual (individuals who are excluded from wide access to ICTs), organizational (refers to
companies that lag behind in terms of ICTs adoption) or global (when some
countries lag behind in terms of ICTs adoption) perspectives.
As specified above, the concept of digital divide refers mostly to the division between societal groups that possess expansible and infinite access to
most of recently developed “knowledge products”2 (see [Adriani and Becchetti, 2003]) and hardware, and those who are excluded from such benefits.
While studying the magnitude of past and present digital divides, the applied
definition plays crucial role. Results of the study can differ significantly
when different notions and measurement methods are implemented. In this
paper we shall imply the reductionist definition of digital divides, assuming
2
Software and databases.
ekonomia 31
97
Ewa Lechman
that it refers to a gap between those who have access to ICTs and are able to
use it, and those who—regardless of the reasons—do not have such an
opportunity.
2. Technology convergence—theoretical outline
As is widely known, the idea of convergence, directly derived from growth
theory, is simple and easy to interpret. The process of convergence reports on
growing cohesion among selected objects (countries in most cases), in terms
of arbitrary assumed variables (indicators), which mainly is assumed to be
national income per capita. It shows negative correlation between GDP per
capita growth rates and initial GDP per capita level (natural logarithm of
GDP). Such a notion of convergence also refers to the catching-up hypothesis
(see [Abramowitz, 1986]) which asserts that being backward in the GDP level
carries a great potential (possibility) of rapid advance. It implies that in the
long run perspective, GDP per capita growth rates are inversely related to the
initial level of GDP or any other economic indicator (if applied). However the
results of convergence process analysis are valuable, they do not explain any
causality between variables, or any other factors that could possibly speed up
or impede the process. In the following paper, we assume that convergence
should be perceived in terms of technology exclusively.
In the paper, we use the idea of unconditional b-convergence, s-convergence and quantile-convergence. Despite being easy in nature, the
estimates of b-convergence have a few recognized limitations. The estimated
coefficients report solely on the central tendency of distribution ignoring the
behavior of a variable in its non-central locations. In such a case, despite
having confirmed—or rejected—the hypothesis of unconditional technology
b-convergence, it gives just a simple idea of an average evolution of variable
growth behavior over time. To draw more detailed conclusion about technology distribution we run additionally q-convergence (quantile convergence),
a methodology based on quantile regression analysis. The q-convergence (see
[Castellacci, 2006 and 2011]), a non-parametric method (see [Koenker et
Bassett, 1978, 2001, 2005], see also [Hao and Naiman, 2007]), provides more
detailed information about the behavior of variable distribution in a set of j
quantiles (percentiles)3. Since any number of quantiles can be applied in the
analysis, it allows modeling arbitrary predetermined position of distribution4. Additionally, this methodology tells a lot about variable behavior in
certain quantiles of distribution including its left and right tails. Using the
q-regression is especially useful when variables’ distributions are skewed.
In addition, we will tests for the s-convergence. The methodology gives
a general idea about dynamics of the variability of the particular variables
3
4
98
The numbers of quantile is set arbitrary by the author.
Hao L., Naiman D.Q., Quantile regression, SAGE Publications 2007.
ekonomia 31
distribution. Based on that we learn about the increase/decrease of the
dispersion of given variables in the studied time span.
Along with the convergence process analysis, there emerges the question
of creating groups of “rich” and “poor” countries. In literature the problem is
recognized as convergence clubs formation (see [Rostow, 1980; Ben David,
1997; Quah, 1993, 1996]). The notion of “convergence club” refers to an identified group of countries where the catching-up hypothesis was positively verified. Consequently, within the group, the growing cohesion (for example in
terms of GDP per capita) can be observed. Baumol [1986], in his study, distinguishes three types of convergence clubs. The first one refers to high income
industrialized countries which are supposed to converge strongly, the second—to middle income countries when the catching-up hypothesis may or
may not be confirmed (in any case, convergence is not supposed to be so
strong as in the high income group), and third—to low income countries,
where convergence is hardly visible. In literature (see [Quah, 1996]), there is
also a distinct classification of convergence clubs. The first one named “upward convergence”—refers to the group of relatively backward countries
which tend to catch-up with the rich ones; the second one is called “downward convergence” and is observed in the group of relatively advanced economies where growth rates (for example GDP per capita) are at very low levels—close to 0% per annum, or even happen to be negative. Note that with
such a distinction, any convergence tendencies within groups do not have to
be reported. It rather explains interactions between distinct country groups.
The term of “club convergence”, along with the issues just discussed, also
refers to the situation when certain economies tend to stay in the same “club”
over time, which means that they hardly improve their relative position, i.e.
country X was classified as poor in 1970, and after a 30-year period is still
classified as such. Such an approach generates the emergence of two theoretical country clusters (groups): poor (“bottom cluster”) and rich ones (“top cluster”). Clearly it does not mean that certain indicator values for countries
within clusters (clubs) do not change. In fact, they do, the changes, however,
are not dynamic and strong enough to let a country move from the bottom to
the top cluster.
3. Data—preliminary analysis
The data set we employ for the analysis consists of 145 countries, for which
we have managed to complete statistical data of 5 different ICTs variables.
The time coverage is 2000–2010. The variables show a country’s achievements
in adoption of most common information technologies tools, and can be
treated as proxies of a country’s development on the given field. The indicators are: Fixed telephone lines5 (FTL) per 100 inhabitants, Fixed internet
5
In the following parts of text, we use abbreviations.
ekonomia 31
99
Ewa Lechman
subscriptions (FIS) per 100 inhabitants, Fixed broadband subscriptions
(FBS) per 100 inhabitants, Internet users (IU) per 100 inhabitants, Mobile cellular subscriptions (MCS) per 100 inhabitants6.
A preliminary descriptive data analysis explains basic characteristics of
selected variables. The country sample is broad (it covers 145 economies) and
allows to detect world general tendencies in information and communication
technologies adoption and usage. Following the descriptive statistical analysis (see Table 1 and Graph 1 below), we estimated densities functions for the 5
variables—in 2000 as the start year and in 2010 as the end year, to check for
changes in world distributions of ICTs.
Table 1.
Summary descriptive statistic and Gini coefficients. Selected ICTs indicators. Years 2000 and
20107, 145 countries
Variable
Mean
Std. Dev.
Min value
Max value
Kurtosis
Gini coeff.
FXTEL2000
23.6
21.9
0.019
86.07
–0.529
0.512
FXTEL2010
22.6
18.7
0.063
82.06
–0.136
0.459
(–1)
(–3.2)
+0.044
(–4.01)
–
(–0.053)
changes in FTL
FXINTER2000
4.71
7.6
0.0037
39.30
5.32
0.718
FXINTER2009
12.0
12.5
0.010
47.35
–0.307
0.557
+7.29
+4.9
+0.0063
+8.05
–
(–0.161)
FXBROAD~2000
1.3
3.12
0
22.58
16.8
0.830
FXBROAD~2010
11.1
12.2
0
63.83
1.18
0.583
changes in FBS
+9.8
+9.08
0
+41.25
–
(–0.247)
INTUSERS2000
10.03
13.7
0.0059
51.3
1.3
0.662
INTUSERS2010
39.7
27.4
0.72
95
–1.13
0.332
changes in IU
+29.67
+13.7
+0.71
43.7
–
(–0.33)
MOBILES~2000
20.2
24.29
0
81.48
0.009
0.618
MOBILES~2010
96.5
39.3
3.526
206.42
–0.038
0.228
+76.3
+15.01
0.3526
124.94
–
(–0.39)
changes in FIS
changes in MCS
Source: own calculations using STATA 11.2 and GRETL Raw data drawn from ITU databases
2011.
The sample consists of 145 world economies. Statistics in Table 1, give
a general idea of the level of adoption of given ICTs in selected countries and
presents preliminary data descriptive analysis results. Additionally we have
estimated the Gini coefficient in 2000 and 2010, to check for changes in distribution inequalities of ICTs variables. The period employed for the analysis is
6
7
100
Detailed definitions of each variable are put in Appendix 1.
For Fixed Internet Subscr. data, the time span is 2000–2009.
ekonomia 31
widely thought as the one when fast changes in ICTs adoption were taking
place worldwide. As seen from statistics values in Table 1, the hypothesis on
fast ICTs diffusion definitely can be confirmed. Also, it is clearly visible, that
these changes happen at a different pace when different ICTs tools are taken
into account. Except for the FTL variable, where we observe hardly any
changes in its adoption, in case of the following four indicators (FIS, FBS, IU,
MCS) the changes are astonishing. The fixed telephone lines are perceived
rather as a kind of traditional means of communication, now being consequently substituted by new ones. That is the reason why we observe minimal
changes in mean and standard deviation values. We can conclude that cross-country level of fixed telephones adoption is stable in analyzed period
2000–2010, as well as its distribution (the Gini in 2000 was 0.51; in 2010—
0.459). Distinct conclusions are drawn when analyzing FIS, FBS, IU and MCS
statistics. In all four cases statistics report on crucial changes, both in absolute levels of ICTs’ implementation and in Gini values. It shows how dynamic
ICTs are being adopted across countries. In each case we observe high increments in mean values (highest changes in case of MCS, change from 20.02 in
the year 2000, to 96.5 in 2010), as well as great increases in Min and Max values for each variable. That proves a fast growth in basic ICTs tools adoption,
not only in high-income countries, but also in middle- and low-income ones.
In addition, such positive changes show that in the period 2000–2010, a great
majority of low and middle-income economies have undergone a kind of
“technological revolution”, and were adopting basic ICTs tools countrywide.
The only exception constitutes the case of FBS, where still in 2010; the Min
value is zero for some countries, which means that they cannot benefit from
the broadband Internet tool8. Apart from great changes in absolute variables’
levels, we also observe substantial changes in Gini coefficients. For all
indicators, the Gini values were much higher in 2000 than in the year 2010
(see Chart 1).
The greatest decrease in Gini coefficient has been noted for IU: a 33 percentage points decrease, and MCS: a 39 percentage points decrease over the
period of 2000–2010. To have an idea about the magnitude of changes in inequalities, see Chart 2 presenting Lorenz curve for MCS variables in 2000, after in 2010—Chart 3.
To learn more about the worldwide distribution of ICTs tools on country
level, we estimate distributional graphs for each variable separately. The following 5 charts (Chart 4, 5, 6, 7, 8) show densities function estimates9. To show
changes in distributions clearly we have prepared two-way charts for each
variable.
8
The countries identified with “0” value of FBS in 2010 are: the Comoros, Iraq, Ethiopia,
Eritrea and Burundi. Data according to ITU database (2012).
9
In each case we apply Gaussian Kernel densities.
ekonomia 31
101
Ewa Lechman
Changes in GINI coeff. FTL, FIS, FBS, IU, MCS. Period 2000–2010
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
GINI2000
GINI2010
0.0
FXTEL
FXINTER
FXBROAD
INTUSERS
MOBILES
Chart 1.
0
G en. Lore nz (MO BILES UBS2000)
5
10
15
20
Changes in Gini coefficients for FTL, FIS, FBS, IU and MCS. Period 2000–2010
Source: own elaboration using STATISTICA 10.0.
0
.2
.4
.6
Cumulative population proportion
.8
1
Chart 2.
Lorenz curve for MCS variables in 2000
Source: own elaboration using STATA 11.2.
102
ekonomia 31
0
G en. Lore nz (MO BILES UBS2010)
20
40
60
80
100
0
.2
.4
.6
Cumulative population proportion
.8
1
Chart 3.
0
.005
.01
.015
.02
.025
Lorenz curve for MCS variables in 2010
0
20
40
60
80
x
kdensity FXTEL2000
kdensity FXTEL2010
Chart 4.
FTL distributions. 2000 and 2010
Source: own estimates applying STATA 11.2.
ekonomia 31
103
0
.05
.1
.15
Ewa Lechman
0
10
20
30
40
50
x
kdensity FXINTER2000
kdensity FXINTER2009
Chart 5.
0
.2
.4
.6
.8
1
FIS distributions. 2000 and 2009
0
20
40
60
x
kdensity FXBROADBAND2000
kdensity FXBROADBAND2010
Chart 6.
FBS distributions. 2000 and 2010
104
ekonomia 31
0
.02
.04
.06
0
20
40
60
80
100
x
kdensity INTUSERS2000
kdensity INTUSERS2010
Chart 7.
0
.005
.01
.015
.02
.025
IU distributions. 2000 and 2010
0
50
100
x
kdensity MOBILESUBS2000
150
200
kdensity MOBILESUBS2010
Chart 8.
MCS distributions. 2000 and 2010
ekonomia 31
105
Ewa Lechman
For FTL, we can hardly observe any changes in distribution. The densities
functions look very similar both for the year 2000 and 2010. Similar conclusions were already drawn from descriptive statistics, as well as we observed
only slight decrease in Gini coefficient. Opposing to that, Charts 5, 6 and 7
show substantial changes in variables (FIS, FBS and IU) distributions. The
density function plots, for the year 2000, show one-peak distribution accompanied by long right tail. It shows highly uneven distribution of ICTs tools in
2000 among countries, but also proves the existence of numerous groups of
countries where the ICTs adoption was at an extremely low level. At the same
time, the distribution of ICTs among middle- and high-income countries was
highly uneven (see long left tail). In 2000, in terms of ICTs adoption, the group
of low-income countries was rather homogenous, while the group of middleand high-income economies was much more diversified. Over the period of
2000–2010, the situation changed significantly. Looking again at the same
charts (see Chart 5, 6, 7), but for densities functions in 2010, we note that line
shapes differ substantially drawing a different picture of the problem. The
densities lines show highly advanced stratification processes of ICTs distribution among countries. Such changes are a consequence of dynamic process
of ICTs implementation across countries, and the disappearance of high left
peak proofs that in the countries ICTs adoption level has increased. The
group of countries experiencing a high level of ICTs deprivation in the year
2000, could enjoy using new technologies at an acceptable level as early as
2010. The ICTs diffusion process, despite having an unquestionable positive
impact, has also led to a great diversification of countries in terms of ICTs
adoption. The sharp division on the world map has disappeared, but in
exchange, countries (as a group) are much more diversified in terms of ICTs
implementation.
The last chart (Chart 8) refers to world distribution of mobile cellular subscribers in the countries included in the sample. In the year 2000, we can observe a clear polarization—see twin-peak density function, on the world map.
Each peak stands for a relatively homogenous group of economies with a similar level of MCS, while the differences between the two groups are high.
A high left peak of distribution stands for low income (and probably low-middle-income) countries with a relatively poor adoption of mobiles in their societies. The right peak of distribution stands for a group of relatively rich countries which enjoy a higher level of mobiles usage. The polarization disappeared in the year 2010, when we observed a sole, centered peak. Such
changes show a great increase in the usage of mobile phones, especially in
low- and medium-income countries.
4. Do countries converge in the field of technology?
As presumed in section 2, we intend to verify the hypothesis of inter-country technology convergence in the time span of 2000–2010. To learn more
about convergence tendencies—or a lack of them—we run a 3-step analysis.
106
ekonomia 31
First, we check for traditional beta-convergence (1-step), then we estimate
quantile—convergence (2-step) and finally sigma-convergence (3-step). Following the idea, in each step, five separate regressions will be estimated. We
assume that the dependent variables are the growth rates of the selected
ICTs indicators in the period 2000–2010, while as explanatory variables are
used the initial levels (in the year 2000) of the respective indicators. Therefore, we limit the analysis to one regressor. The data and time coverage is
analogous to section 2.
a. The b-convergence testing—1-step.
As assumed, each regression shall have just one regressor—the initial
level10 of a given variable in a given country. We estimate 5 different equations, for each indicator separately. The models 1(a), 2(a), 3(a), 4(a) and 5(a)
are identifiable as following:
Yj(FTL2000–2010) = a + bj(ln_FTL2000) + ej
(1a)
Yj(FIS2000–2010) = a + bj(ln_FIS2000) + ej
(2a)
Yj(IU2000–2010) = a + bj(ln_IU2000) + ej
(3a)
Yj(FBS2002–2010) = a + bj(ln_FBS2002) + ej11
(4a)
Yj(MCS2000–2010) = a + bj(ln_MCS2000) + ej
(5a)
Where, Yj denotes the average annual growth rate of a given technology indicator in j-country. The b-coefficient reported in a set of regression is crucial
to verify the hypothesis on existence the convergence among the set of
countries. If the b-coefficients result to be negative and statistically significant, it suggests that countries tend to converge. Complete analysis results
are presented in Table 2 (see below).
Table 2.
b-convergence estimation results. ICTs variables, time coverage 2000–2010
_cons
b-coeff.
FTL
6.33
12
–1.96
(–10.57)13
0.438
FIS
15.89
–2.99
(–7.96)
0.307
Variable
10
11
12
13
R-squared
In the year 2000.
Estimates for 108 countries.
0.05 significance level.
t-statistics in parenthesis.
ekonomia 31
107
Ewa Lechman
Variable
_cons
b-coeff.
R-squared
IU
28.43
–5.43
(–22.24)
0.775
FBS
33.28
–7.55
(–20.62)
0.80
MCS
41.29
–8.14
(–46.86)
0.93
Source: own estimations using STATA 11.2.
In the equations (1a), (2a), (3a), (4a) and (5a), the estimated parameters result to be negative and statistically significant14 in each case. The negative b
parameter, let us to confirm the hypothesis on existence of unconditional
technology convergence among the 145 countries applied for the study. In the
case of FTL, the coefficient results to be the lowest, however still negative.
The regression (1a) refers to the fixed telephone lines it is rather not surprising that its adoption does not play a crucial role in the economy. In 63 countries out of the 145, growth rates presenting changes in per inhabitant fixed
lines are negative. This proves a substitution of traditional means of communication by modern ones. In this case, we would conclude on substitution of
fixed line by mobile phones.
In regressions (2a), (3a), (4a) and (5a) the b-coefficients are still negative
and relatively high. It reports on dynamic unconditional technology convergence process in the analyzed countries. The best score we obtained was the
one of MCS indicator. The coefficient at (–8.14) together with the very high
negative correlation coefficient (–0.96)15 show that process of mobiles phones
implementation is very dynamic. A similar conclusion can be drawn from
Chart 8 (see previous section). In terms of per inhabitant, an average usage of
mobile phones has grown enormously, both in low- and high-income economies.
It is no surprise that countries that had a relatively low level of ICTs adoption in the year 2000, tended to grow at an enormously high pace in the period
of 2000–2010. Thanks to that effect they have an opportunity to get closer to
economies already advanced in ICTs usage. The results also report on catching-up effect in terms of new information and communication technologies
application and usage in the 145 economies. However, the effect is positive
and can influence enormously the socio-economic development path in lowand middle-income countries, it shall be underlined that these economies do
not create new technologies. They just adopt them at a relatively low cost.
ICTs implementation also enhances higher investments in basic human skills
enabling to use these technologies effectively. The so-called “digital literacy”
or “digital readiness” is a prerequisite to get gains from ICTs usage.
14
15
108
For each equation the p-value < 0.05.
Own calculations using STATA 11.2.
ekonomia 31
b. The q-convergence testing—step 2.
In the following subsection, we run a set of quantile regressions for each of
the ICTs indicators. Applying the non-parametric method will help us find
out more on the variables’ behavior in non-central locations of the respective
distributions. We use a set of mathematical formulas to estimate technology
convergence—if reported—on arbitrary assumed quantiles.
Yji(FTL2000–2010) = a + bji(ln_FTL2000) + ej
(1b)
Yji(FIS2000–2010) = a + bji(ln_FIS2000) + ej
(2b)
Yji(IU2000–2010) = a + bji(ln_IU2000) + ej
(3b)
Yji(FBS2002–2010) = a + bji(ln_FBS2002) + ej16
(4b)
Yji(MCS2000–2010) = a + bji(ln_MCS2000) + ej
(5b)
The i stands for an ith quantile of the growth distribution of the indicator. The
author arbitrarily assumes the estimations of 20th, 40th, 60th and 80th quantile
of the respective ICTs indicators distribution. As in previous cases, the regressions consist of one predictor variable. The results of the quantile regressions are presented in Table 3 (see below).
Table 3.
Fixed Telephone Lines, Fixed Internet Subscribers, Fixed Broadband Subscribers17, Internet
Users, Mobile Cellular Subscribers. The q-convergence estimates. 145 countries. Years 2000—2010
q-convergence (the (coefficients)
20th quantile18
40th quantile
60th quantile
80th quantile
FTL
–1.28
(–5.10)19
–1.73
(–8.79)
–2.06
(–10.18)
–2.52
(–18.37)
FIS
–1.85
(–3.82)
–2.25
(–7.04)
–3.47
(–17.30)
–5.20
(–16.56)
IU
–4.24
(–13.73)
–5.22
(–30.05)
–6.29
(–38.79)
–6.95
(–38.52)
FBS20
–5.73
(–1.34)
–6.98
(–24.49)
–8.07
(–26.56)
–9.36
(26.75)
MCS
–7.71
(–41.37)
–8.38
(–50.06)
–8.63
(–57.61)
–9.03
(–47.71)
Indicator
Source: own estimations using STATA 11.2.
16
17
18
19
20
For the MCS the regressions are run for 99 economies in the period of 2002–2010.
The estimates for the sequent quantiles are always run in the whole country sample.
The t-statistics are put in parenthesis.
ekonomia 31
109
Ewa Lechman
The quantile regression analysis completes the unconditional beta-convergence, and shades more light on the dynamic of inter-country technology
convergence. In Table 3, q-regression coefficients are reported on the 20th,
40th, 60th and 80th quantiles for each ICTs variable separately. In each case,
the regression coefficients are the lowest in the first (20th) quantile, and are
increasing in the following 3 quantiles, reaching the highest level in the 4th.
For FTL, FIS, FBS, IU and MCS, the coefficients turn out to be higher in the
4th quantile than in case of the inclusion if the whole distribution. That is because the 4th quantile’s estimate does not include long right tail of the variables’ distributions.
The overall results show clearly that in countries with a relative low initial level of ICTs adoption, the elasticity of ICTs implementation is also relatively lower. That suggests poorer ability of underdeveloped countries to acquire and use new ICTs tools. This is probably due to a relatively low cost of
mobiles’ adoption and a great ability to use it with no special human skills
requirements.
c. The s-convergence testing—3-step
Thirdly, we turn attention to the sigma-convergence testing, which reports
on an increase or decrease in the coefficients of variation of certain ICTs
variables. Such an approach shows the general tendency in growing or diminishing diversification within an analyzed group of countries in terms of dispersion of given variables’ distribution.
Here below, we present results of sigma-convergence estimates (see Table
4 below).
Table 4.
Sigma-convergence coefficients estimates for FTL, FIS, FBS, IU and MCS. Years 2000 and
2010
Variable
Coeff. of variation in
2000
FTL
93.0
FIS
Coeff. of variation in
2010
% change in variation coefficients levels
in the period 2000–2010.
82.93
(–10.92%)
162.91
103.85
(–36.25%)
IU
229.80
110.42
(–51.95%)
FBS
137.08
69.04
(–49.63%)
MCS
120.16
40.74
(–66.09%)
Source: own calculations using STATISTICA 11.2, based on data from ITU 2012.
As expected, also sigma-convergence tests show enormous changes in
variation coefficients for selected ICTs indicators. The greatest decrease in
coefficients of variation is observed in case of Internet users (decrease of almost 52%) and—again—mobile phones subscribers (decrease of 66%). Provided such results we can again strongly confirm that in the period of 2000–
110
ekonomia 31
–2010, a fast and dynamic process of ICTs diffusion across countries took
place.
To sum up, in the fourth section we have tested for convergence process in
145 economies in the time span of 2000–2010. For the convergence analysis,
we have chosen three methods: b-convergence, s-convergence and q-convergence. As proxies of ICTs adoption level we have selected 5 indicators: fixed
telephone lines, fixed Internet subscribers, Internet users, fixed broadband
subscribers and mobile cellular subscribers. Given statistics draw a clear
picture of overall basic ICTs tools usage in each of 145 countries. General
results from convergence testing—regardless of methodology—are similar
and prove a strong and fast inter-country technology convergence. This is
mainly due to fast ICTs adoption, especially in low- and middle-income
countries. However, the process of cross-country ICTs adoption is positive
and generates great possibilities for ICTs users, it shall be underlined, that in
a great number of countries the average use of basic ICTs is still relatively
low. In addition, it should be noted that fast technology convergence does not
imply directly that the technology gaps will disappear. This is a long-term
process and requires huge financial resources and great improvements in
basic human skills, so that the ICTs adoption would be effective and gains
generating.
The gap still stays, which can be easily concluded from most recent ICTs
cross-country adoption statistics. We need to remember that ICTs implementation and usage is also growing rapidly in high and medium income economies. The process is not static—quite the contrary—is it characterized as
highly dynamic in each country and from a global perspective.
5. And what about technology club convergence?
As stated in the first section, the objectives of the paper are twofold.
Firstly, we checked for catching-up (determined by technology convergence)
effects in the group of selected 145 countries (which has been confirmed), and
secondly, we aimed to identify the convergence clubs formation within the
same group of economies. Following the Schumpeterian21 model of convergence clubs, we assume that countries differ significantly from one another.
These differences cover inter alia: per capita income level, GDP per capita
growth dynamics, basic human skills, absorptive capacity of human capital,
ability to absorb and adapt innovations and new technologies. We also make
an assumption that low- and middle-income countries (relatively backward
economies) have poor absorptive capacity which enables them to jump from
the “poor club” into the “rich club”. The overall country ability to adapt and
use new technologies is a prerequisite to change the club.
21
Kang S.J. [2002].
ekonomia 31
111
Ewa Lechman
To group countries, we employ a country's dynamics based classification
approach, which stands for classifying countries according to the magnitude
of progress they made in the period of 2000–2010. To shed more light on the
idea of country clustering, we present a theoretic scheme of clubs (see Chart
9 below). Cluster I (club I) includes countries which are mostly highly developed in terms of ICTs adoption (in year (1) and (2) these countries enjoyed the
highest level of development); Cluster II (club II)—countries that in the examined time span managed to change their relative position from low to high
developed countries; Cluster III (club III)—countries relatively backward in
terms of ICTs adoption, countries which did not manage to jump into the
“rich club”; Cluster IV (club IV)—a group of countries that worsened their
relative position in the assumed time span.
Cluster I
Cluster III
Cluster IV
LN(x) year 2
Cluster II
LN(x) year 1
Chart 9.
Convergence clubs (clusters)—theoretical framework
Source: own elaboration.
To check for the club convergence, we plot 5 ICTs variables separately
(see Charts 10, 11, 12, 13 and 14). In each we divide coordinate system into 4
parts, pointing 4 clusters (see Chart 10. for details). We draw the vertical line
at value “0” on the axis LN(x)year1, to make a clear division between Cluster
III and IV. The zero value at the LN(x)year1 axis indicates the value of an indicator for a country in 2000 at about 1 (units). In this case, the initial value
“1” for a given indicator—in the year 2000, a threshold is assumed for an initial classification of poor and rich countries. We have named the following
112
ekonomia 31
clubs as: Cluster I—advanced countries, Cluster II—fast followers, Cluster
III—lagging behind countries, Cluster IV—marginalized countries.
First, we check for club convergence in the case of fixed telephone lines
(see Chart 10).
FTL 2000–2010. Club convergence
5
4
3
LN Tele 2010
2
1
0
–1
–2
–3
–4
–5
–4
–3
–2
–1
0
1
2
3
4
5
LN Tele 2000
Chart 10.
Convergence club for FTL. 2000–2010
Most of the 145 countries belong to Cluster I—highly developed economies
in terms of fixed telephones adoption. Only 8 economies (see Table 5.) managed to jump from the poor into the rich club (see Cluster II), by moving from
the third quarter of the coordinate system into the second one. Very few countries still stay in Cluster III, which means that they are still lagging behind in
terms of FTL.
The second plot (Chart 11), shows club convergence for FIS indicator. In the
case of 42 economies (the list of economies is specified in Table 5., see below)
belong to Cluster II—these are fast following countries that in the period of
2000–201 managed to change their position in the world ranking. However, still
many countries stayed in the lagging behind group. It proves that in these countries the process of fixed Internet adoption was not dynamic enough to be classified as a member of Cluster II. The average per 100 inhabitants fixed Internet
implementation in countries from Cluster III, although slightly higher than in
the year 2000, in 2010 was still at a very low level—below 122 in each country.
22
1 per 100 inhabitants.
ekonomia 31
113
Ewa Lechman
FIS 2000–2009. Club convergence
5
4
3
Fixed INT LN 2009
2
1
0
–1
–2
–3
–4
–5
–6
–4
–2
0
2
4
6
Fixed INT LN 2000
Chart 11.
Convergence club for FIS. 2000–2009
5
4
3
LNFBS 2010
2
1
0
–1
–2
–3
–10
–8
–6
–4
–2
0
2
4
LNFBS2002
Chart 12.
Convergence club for FBS. 2002–2010
114
ekonomia 31
Quite a similar situation is shown in Chart 12, presenting club formation
when fixed broadband (FBS) subscribers are taken into account. In Cluster II
we find 57 countries (fast followers) which is the best score out of the 5 cases
analyzed. It should be underlined that all 57 economies, in the year 2002 were
classified as poor in terms of FBS. In the year 2002 the average per 100 inhabitants fixed broadband adoption level was considerably below 1 per 100 inhabitants. By contrast, in 2010, each of the countries enjoyed a significantly
higher level of FBS adoption. Still, the group of countries (Cluster II) is highly
diversified. Although there are many countries where the FBS adoption level
stands at about 30–40 units per 100 inhabitants23, in many economies the analogous values are just above zero. Hopefully, in the case of FBS, Cluster III is
poorly populated and no country is classified as a marginalized economy.
IU 2000–2010. Club convergence
5
4
INT Users LN 2010
3
2
1
0
–1
–6
–4
–2
0
2
4
6
INT Users LN 2000
Chart 13.
Club convergence for IU. 2000–2010
When analyzing the Internet user (IU) indicator, we have found out
a highly positive situation. Many countries are classified as rich (Cluster I),
and in the period of 2000–2010 another 37 countries managed to join the rich
group. Unfortunately, which is highly undesirable in the analyzed case, we
have observed that two economies (Congo and Ethiopia) were classified as
lagging behind countries (Cluster III).
23
The highest value was noted for Liechtenstein—63.8.
ekonomia 31
115
Ewa Lechman
MCS 2000–2010. Club convergence
5.5
5.0
4.5
Mobile LN 2010
4.0
3.5
3.0
2.5
2.0
1.5
1.0
–5
–4
–3
–2
–1
0
1
2
3
4
5
Mobile LN 2000
Chart 14.
Convergence club for MCS. 2000–2010
Chart 14 (see above) pictures a slightly different situation than in the previous cases. The group constituting Cluster II is still quite numerous (35 countries), and no countries have been classified as lagging behind and/or
marginalized. The “construction” of Cluster I, however, is extraordinary.
There are many countries that were classified as very poor in the year 2000,
and achieved the level of MCS indicator of highly developed economies in
2010. This proves that process of mobile phones diffusion was very dynamic
in the period of 2000–2010. It should be born in mind that similar conclusions
were drawn from descriptive statistics analysis and then from convergence
process analysis. In the period of 2000–2010, the average mobile phones subscribers level increased from 20.2 to 96.5, while the maximum level grew from
81.48 to 206.6224. Cluster I was significantly diversified internally. Along with
the highly developed countries such as Germany or Sweden, there are economies like Swaziland, Togo, Senegal or Belize, traditionally classified as low
developed countries. Such fast changes are possible mainly due to a very low
cost of mobile phone adoption in a society, and relatively low human skill requirements to use them effectively. This confirms again the hypothesis on the
catching-up process taking place especially in low developed economies.
24
116
Always in terms of per 100 inhabitants.
ekonomia 31
Table 5.
Members (countries) of Cluster II for FTL, FIS, FBS, IU, MCS
FTL
FIS
FBS
FBS cont.
IU
MCS
Angola
Albania
Argentina
Mexico
Albania
Albania
Cambodia
Angola
Armenia
Moldova
Angola
Angola
Eritrea
Armenia
Azerbaijan
Mongolia
Azerbaijan
Armenia
Ethiopia
Azerbaijan
Bahrain
Morocco
Bangladesh
Bangladesh
Lao Rep.
Belarus
Bahrain
New Caledonia
Benin
Belarus
Malawi
Bolivia
Belarus
Oman
Bhutan
Benin
Mauritius
Bosnia and Herz.
Bolivia
Panama
Burkina Faso
Bhutan
Togo
Bulgaria
Bosnia and Herz.
Peru
Burundi
Burkina Faso
Cape Verde
Brazil
Philippines
Cambodia
Burundi
ekonomia 31
China
Brunei
Poland
Djibouti
The Comoros
Colombia
China
Puerto Rico
Egypt
Congo
Costa Rica
Colombia
Qatar
Erithrea
Djibouti
Djibouti
Costa Rica
Romania
Georgia
Eritrea
Dominican Rep.
Cyprus
Russia
Ghana
Ethiopia
Ecuador
Czech Rep.
Saudi Arabia
Indonesia
Ghana
Fiji
Ecuador
Slovak Rep.
Iraq
India
Georgia
Egypt
South Africa
Kenya
Iraq
India
Faroe Islands
Sri Lanka
Lao RP
Kenya
Jordan
French Polynesia
Surinam
Madagascar
Kyrgyzstan
Maldives
Georgia
Thailand
Malawi
Lao Rep.
Moldova
Grenada
Tonga
Mauritania
Madagascar
Mongolia
Ireland
Trinidad & Tobago
Morocco
Malawi
Morocco
Jamaica
Tunisia
Nepal
Mauritania
Namibia
Jordan
Turkey
Nigeria
Nepal
Pakistan
Kuwait
United E.A.
Paraguay
Nigeria
Paraguay
Latvia
Venezuela
Rwanda
Pakistan
Peru
Lebanon
Senegal
Rwanda
Philippines
Lithuania
Sri Lanka
Syrian Rep.
Russia
Malaysia
Tanzania
Rwanda
Maldives
Tonga
Salvador
Mauritius
Uganda
Sri Lanka
Uzbekistan
Surinam
Vanuatu
Swaziland
Yemen
117
Ewa Lechman
FTL
FIS
FBS
FBS cont.
IU
MCS
Syrian Rep
Tanzania
Thailand
Tunisia
Ukraine
Uzbekistan
Vanuatu
Yemen
Source: own elaboration based analysis results from section 5.
6. Final remarks
The main scope of the study was to examine cross-national disparities in
the field of new information and communication technologies adoption and
usage. In order to achieve this aim, we have run basic descriptive statistical
analysis (Table 1.), checked for changes in five different ICTs tools worldwide
distributions (Charts 3–7), confirmed the hypothesis of the catching-up process taking place (applying beta, sigma and quantile convergences approach),
and finally, we have checked for convergence clubs formation in the assumed
country sample. The general conclusions, drawn on the basis of the 145-country sample in the period of 2000–2010, are the following:
a. In most countries the process of ICTs diffusion is fast and dynamic.
b. With regard to 4 ICTs indicators a huge increase in theirs average per 100
inhabitants adoption level has been observed (except for fixed telephone
lines, where slight changes occurred).
c. In the year 2000, the characteristic twin-peak shape distribution line was
observed, which proved the existence of two homogenous groups of countries that differed significantly in terms of ICTs adoption. Reversely, in
2010, the twin-peak curve disappeared and in the global ICTs distribution
we can observe stratification, rather than polarization, tendencies. In the
year 2010 the group of 145 countries was much more diversified in terms of
ICTs adoption than in 2000.
d. Also, substantial decrease in Gini coefficients for all five technology indicators took place. Which provides evidence of the fact that along with the
process of fast ICTs tools diffusion across countries, the inequalities in
their implementation are declining, which is taken to be a very positive
phenomenon.
e. The greatest changes in ICTs adoption and usage have been observed in
the group of relatively low income countries. Many backward economies
managed to make a huge step forward in the new technologies implementation. However, there still remains quite a numerous group of countries
118
ekonomia 31
which find themselves in a very unfavorable position, and are still lagging
behind in terms of ICTs implementation.
f. Analysis results also show a dynamic technology convergence among
countries—regardless of the methodology applied. If so, the catching-up
process occurs at the same time.
g. We have managed to identify different technology convergence clubs
(clusters). In the case of each ICTs indicator, there are many countries belonging to Cluster II which constitutes a group of countries that were classified in the year 2000 as underdeveloped25, where the ICTs adoption
growth rates were higher than in the highly developed countries. The extraordinary growth dynamics let them to catch-up with the developed
economies, and in the year 2010 they achieved a level of ICTs adoption
comparable with that of the highly developed economies.
h. Still, in the case of all 5 indicators, there are a few economies in Cluster
III—those are countries which were permanently lagging behind and did
not manage to catch up with highly developed economies in the period of
2000–2010.
i. Fortunately, only in the case of Internet users (IU) four countries were
classified as Cluster IV economies, constituting a club of marginalized
countries.
Looking at the issues discussed from a broader, global perspective, the
convergence process in terms of ICTs adoption can be easily observed. That
leads to the simple conclusion the low-income countries which are also the
ones with an initial low ICTs implementation, have a great ability to catch-up
with highly developed ones, which is mainly due to the unique ability of ICTs
to spread at a high pace, and at low cost at the same time. In the period of
2000–2010 quite a number of underdeveloped countries managed to change
their position in world ranking, achieving levels of ICTs adoption comparable to the ones we have observed in highly developed economies.
References
Abramowitz M., 1986, “Catching up, Forging Ahead, and Falling Behind,” Journal of
Economic History Vol. 46, Issue 2, p. 385–406.
Adriani F., Becchetti L., 2003, “Does Digital Divide Matter? The Role of Information
and Communication Technology in Cross-Country Level and Growth Estimates?,”
CEIS Tor Vergata—Research Paper Series Vol. 2, No. 4.
Aghion P., Durlauf S.N., 2008, Handbook of Economic Growth Volume 1A, Elsevier.
Bagchi K., 2005, “Factors Contributing to Global Digital Divide: Some Empirical Results,” Journal of Global Information Technology Management 8(1), p. 77–83.
Barro J.R., Sala-i-Martin X., 1997, Economic Growth, The MIT Press, 2004.
Baumol W.J., 1986, “Productivity Growth, Convergence and Welfare: What the Long
Run Data Show,” American Economic Review 76(5), p. 1072–1085.
Ben-David D., 1997, Convergence Clubs and Diverging Economies, NBER and CEPR.
Berenger V., Verdier-Chouchane A., 2007, “Multidementional Measures of well-being: Standard of Living and Quality of Life Across Countries,” World Development
Vol. 35, Elsevier.
ekonomia 31
119
Ewa Lechman
Bernard A.B., Jones C.I., 1996, “Technology and Convergence,” The Economic Journal
Vol. 106, Issue 437, p. 1037–1044.
Billon M., Lera-Lopez F., Marco R., 2010, “Differences in Digitalization Levels:
A Multivariate Analysis Studying the Global Digital Divide,” Review of World
Economy 146,p. 39–73.
Castellacci F., 2011, “Closing the Technology Gap?,” Review of Development Economics 15(1), p. 189–197.
Castellacci F., 2006, “Convergence and Divergence Among Technology Clubs,”
DRUID Working Paper No. 06–21
Castellacci F., 2008, “Technology Clubs, Technology Gaps and Growth Trajectories,”
Structural Changes and Economic Dynamics 19 (2008), p. 310–314.
Comin A.D., Eastely W., Gong E., 2008, “Was the Wealth of Nations Determined in 1000
B.C.?,” Working Paper 09–052, Harvard Business School.
Corrocher N., Ordanini A., 2002, “Measuring the Digital Divide: A Framework for the
Analysis of Cross-Country Differences,” Journal of Information Technology 17(1),
p. 9–19.
Davis L., Owen A., Videras J., 2008, “Do All Countries Follow the Same Growth Process?,” see: http://ideas.repec.org/p/pra/mprapa/11589.html
Desdoigts A., 2004, “Neoclassical Convergence Versus Technological Catch-Up:
A Contribution for Reaching a Consensus,” Problems and Perspectives in Management 3/2004.
Dewan S., Riggins F.J., 2005, “The Digital Divide: Current and Future Reserach Directions”, Journal of the Association for Information Systems Vol. 6, No. 12., p.
298–337.
Devaraj S., Kohli R., 2003, “Performance Impacts of Information Technology: Is Actual Usage the Missing Link?,” Management Science 49, p. 273–289.
Dijk van J.A.G.M., 2005, The Deepening Divide: Inequality in the Information Society,
Sage Publications Inc.
Fei J.C.H., Ranis G., 1999, Growth and Development from Evolutionary Perspective,
Blackwell Publishers.
Gancia G., Zilibotti F., 2009, “Technological Change and the Wealth of Nations,” Annual Review of Economics Vol. 1.
Grandville de la O., 2009, Economic Growth. A Unified Approach, Cambridge University Press.
Goldfarb A., Prince J., 2008, “Internet Adoption and Usage Patterns Are Different: Implications for the Digital Divide,” Information Economics and Policy 20, p. 2–15.
Hao L., Naiman D.Q., 2007, Quantile Regression, SAGE Publications.
Haslam P.A., Schafer J., Beaudet P., 2009, Introduction to International Development,
Oxford Univeristy Press.
Hewitt de Alcantara C., 2001, The Development Divide in a Digital Age. An Issues Paper,
United Nations Research Institute for Social Development.
Howard P.N., 2007, “Testing the Leap-Frog Hypothesis. The Impact of Existing Infrastructure and Telecommunications Policy on the Global Digital Divide,” Information, Communication & Society Vol. 10, No. 2, p. 133–157.
James J., 2003, Bridging the Global Digital Divide, Edward Elgar.
Kang S.J., 2002, “Relative Backwardness and Technology Catching up with Scale Effects,” Journal of Evolutionary Economics 12, p. 425–441.
Kauffman R.J., Techatassanasoontorn A.A., 2005, “Is There a Global Digital Divide for
Wireless Phone Technologies?,” Journal of the Association for Information Systems
Vol. 6, No. 12, p. 338–382.
120
ekonomia 31
Kolenda M., 2006, Taksonomia numeryczna, Wyd. Akademii Ekonomicznej we Wroc³awiu.
Koop G., 2009, Analysis of Economic Data, Wiley & Sons.
Kubielas A.S., 2009, Innowacje i luka technologiczna w gospodarce opartej na wiedzy.
Strukturalne i makroekonomiczne uwarunkowania, Wydawnictwo Uniwersytetu
Warszawskiego, Warszawa.
Maurseth P.B., 2001, “Convergence, Geography and Technology,” Structural Change
and Economic Dynamics 12, p. 247–276.
Meier G.M., Rauch J.E., 2005, Leading Issues in Economic Development, Oxford University Press.
Mookherjee D., Ray D., 2001, Readings in the Theory of Economic Development,
Blackwell Publishers.
Neumayer E., 2003, “Beyond income: Convergence in Living Standards, Big Time,”
Structural changes and Economic Dynamics 14.
Ocampo J.A., 2007, Growth Divergences. Explaining Differences in Economic Performance, Zed Books.
Panek T., 2009, Statystyczne metody wielowymiarowej analizy porównawczej, Szko³a
G³ówna Handlowa.
Quah D.T., 1993, “Galton’s Fallacy and Tests of the Convergence Hypothesis,” Scandinavian Journal of Economics 95, p. 427–443.
Quah D.T., 1996, “Empirics for Economic Growth and Convergence,” European Economic Review 40, p. 1353–1375.
Quah D.T., 1996, “Galton’s Fallacy and Tests of the Convergence Hypothesis,” Scandinavian Journal of Economics 95, p. 427–443.
Ranis G., Stewart G., 2005, “Dynamic Links Between The Economy and Human Development,” DESA Working Paper No. 8.
Ray D., 1998, Development Economics, Princeton University Press.
Redding S., Schott P.K., 2003, “Distance, Skill Deepening And Development: Will Peripheral Countries Ever Get Rich?,” Journal of Development Economics 72 (2003),
p. 515–541.
Rostow W.W., 1980, Why the Poor Get Richer and Rich Slow Down?, Austin University
of Texas Press.
Schiller H.I., 1996, Information Inequality: The Deepening Social Crisis in America,
Routledge, NY.
Schramm W.L., 1964, Mass Media and National Development: The Role of Information
in Developing Countries, Stanford University Press.
Seligson M.A., 2008, Development and Underdevelopment, Lynn Rienner Publishers.
Servon L.J., 2002, Bridging the Digital Divide, Blackwell Publishers.
Spanning the Digital Divide. Understanding and Tackling the Issues. A report by briges.
org, 2004.
Stilianos A., Tomkins J., 2010, “Technology Adoption and Club Convergence,” MPRA
21260, March.
Stoneman P., 2002, The Economics of Technological Diffusion, Blackwell Publishers.
Sussman G., Lent J.A., 1991, Transnational Communication: Wiring the Third World,
Sage Publications.
Thirlwall A.P., 2006, Growth and Development with Special Reference to Developing
Economies, Palgrave.
Todaro M.P., Smith S.C., 2009, Economic Development, Pearson Education.
Where Is the Wealth of Nations? Measuring Capital do 21st century, 2006, The World
Bank, Washington D.C.
ekonomia 31
121
Ewa Lechman
Vehovar V., Sichel P., 2006, “Methodological Challenges of Digital Divide Measurement,” The Information Society 22
Verbeek M., 2012, A Guide to Modern Econometrics, Wiley & Sons.
Wolff E.N., 2009, Poverty and Income Distribution, Wiley-Blackwell.
WoŸniak M.G., 2008, Wzrost gospodarczy. Podstawy teoretyczne, Wyd. Uniwersytetu
Ekonomicznego w Krakowie.
Yusuf S., Deaton A., Dervis K., Easterly W., Ito T., Stiglitz J., 2009, Development Economics Through Decades, The World Bank.
Zhu K., Kraemer K.L., 2005, “Post-adoption Variations in Usage and Value of E-Business by Organizations: Cross-Country Evidence from the Retail Industry,” Information Systems Research 16, p. 61–84.
Appendix 1
Table A1.
Information and Communication Technology indicators
Indicator
Definition
Source
Fixed telephone lines per 100
inhabitants
Fixed telephone lines refer to telephone lines connecting a
subscriber’s terminal equipment to the public switched
telephone network (PSTN) and which have a dedicated port on
a telephone exchange. This term is synonymous with the
terms “main station” and “Direct Exchange Line” (DEL) that
are commonly used in telecommunication documents. It may
not be the same as an access line or a subscriber. The
number of ISDN channels, public payphones and fixed
wireless subscribers are included.
Core ICT Indicators 2010, ITU
Fixed Internet subscribers per
100 inhabitants
Fixed Internet subscribers refer to the total number of
Internet subscribers with fixed access, which includes dial-up
and total fixed broadband subscribers: cable modem, DSL
Internet subscribers, other fixed broadband and leased line
Internet subscribers.
Fixed broadband Internet
subscribers per 100
inhabitants
Fixed broadband Internet subscribers refer to entities (e.g.
businesses, individuals) subscribing to paid high-speed
access to the public Internet (a TCP/IP connection). High
speed access is defined as being at least 256 kbit/s, in one or
both directions. Fixed broadband Internet includes cable
modem, DSL, fibre and other fixed broadband technology
(such as satellite broadband Internet, Ethernet LANs, fixed
wireless access, Wireless Local Area Network and WiMAX).
Subscribers to data communications access (including the
Internet) via mobile cellular networks are excluded.
Internet users per 100
inhabitants
Internet users are those who use the Internet from any
location. The Internet is defined as a world-wide public
computer network that provides access to a number of
communication services including the World Wide Web and
carries email, news, entertainment and data files. Internet
access may be via a computer, Internet-enabled mobile
phone, digital TV, games machine etc. Location of use can
refer to any location, including work.
Telecommunication Indicator
Handbook, ITU
122
ekonomia 31
Indicator
Mobile cellular telephone
subscriptions per 100
inhabitants
Definition
Source
Mobile cellular telephone subscriptions refer to subscriptions
of portable telephones to a public mobile telephone service
using cellular technology, which provides access to the PSTN.
This includes analogue and digital cellular systems, including
IMT–2000 (Third Generation, 3G). Both postpaid and prepaid
subscriptions are included. Prepaid subscriptions are those
where accounts have been used within a reasonable period of
time (e.g. 3 months). Inactive subscriptions, that is, prepaid
cards where a call has not been made or received within the
last 3 months, are excluded.
Source: compilation based on Core ICT Indicators 2010, and Telecommunication Indicator
Handbook, ITU.
Abstract
ekonomia 31
Technology convergence and digital divides. A country-level evidence for the
period 2000–2010
The paper, mostly empirical in nature, investigates issues on cross-national
new information and communication technologies (ICTs) adoption patterns
and growth directions.
In the period of 2000–2010, a great number of countries underwent substantial
changes on the field of ICTs implementation. Many of them made a great
“jump” starting with almost “zero level” of ICTs adoption in the year 2000 and
during the ten- year period were implementing ICTs at an astonishingly high
pace. Despite the obvious positive impact that ICTs have on overall society and
economy condition, rapid changes can also generate higher inequalities on the
field. The paper focuses mainly on capturing these changes. It also aims to confirm or reject the hypothesis on growing inter-country inequalities in ICTs
adoption.
The target of the paper is twofold. Firstly, we explain the magnitude of past and
present differences in digitalization level among countries; secondly, we concentrate digital technology convergence. We apply three approaches to convergence—b-convergence, s-convergence and quantile-convergence (q-convergence), to check if relative division between countries was growing or diminishing in the time span of 2000 to 2010. Additionally, we check if countries of a given
sample tend to form convergence clubs in the relevant years.
The analysis is run for the sample consisted of 145 economies and the time coverage is 2000–2010. All data applied in the research is drawn from the International Telecommunication Union statistical databases [see www.itu.int].
Key words: technology, convergence, ICTs, quantile convergence, clusters,
technology clubs
JEL codes: C22, O11, O50, 033
123

Technology convergence and digital divides.

Transkrypt

Podobne dokumenty

Saturday, 08.08.2015

FIS Calendar Conference in Varna (BUL) from 3rd to 6th June 2015